城市蔬菜产品品种污染监控系统的研究及其应用

:10021566(2005)030001061,1,2,1(11,,51064221,,518040):,,200314,Hotelling,SASDelphi,:;;:O212:ATheresearchoncityvegetableproductvarietyPollutioncontrolsystemanditsaplicationZHANGGuo2quan1,XIAOLi1,ZHANGBing2,CENGuan2jun1(11CollegeofScience,SouthChinaAgriculturalUniversity,Guangzhou,510642,China21SafetyfoodQualitySuperviseInspectionStation,ShenZhen,518040,China)Abstract:Inordertohelpcitizensknowthequalitystatusandsecuritylevelofdailyvegetableproductvariety,MultivariatestatisticalmethodssuchasprincipalcomponentsandHotellingT2,combinedwithSASandDelphi,areappliedtoestablishthepollutioncontrolsystemofcityvegetableproduct.ThesemethodswereusedinthesamplingdataduringJanuarytoApril2003fromBuJiandFuTianfarmproducewholesalemarketinShenZhenCity.Thepollutionindexcontrolsystemcanproviedtimelyinformationaboutvegetableproductqualitystatusandpollutionlevel,controlthevegetableuqality,guaranteecitizenshealthandleadmarketconsumption.Keywords:pollutionindex;multivariatestatistic;controlchart,,,,,,,200314,,,SASDelphi:20031171:©1995-2005TsinghuaTongfangOpticalDiscCo.,Ltd.Allrightsreserved.,,,211(VARIETYINDEX,VI),,212VI(t):,,,::X1=100%:X2=100%(1)X1,X2,,(1)X=X1(t)X2(t)T,X(t),n,t=1,2,,nf[X(t)],,X1X2,()()(),,,,,,X=(X1,X2)N(u,r)p=2,E(X)=U,D(X)=r:Y1=1X=11X1+21X2Y2=2X=12X1+22X2Var(Yi)=Max=1,j=0(j=1,,i-1)Var(X),Yi=iXXi(i=1,2)i(t)Xn2,X(t)ii(t)Y1(t)X,Y2(t)X224320055©1995-2005TsinghuaTongfangOpticalDiscCo.,Ltd.Allrightsreserved.1Yi(t)¦XT(t)i(t)+(t)(2),i=1,2,Y1(t)Y2(t),t=1,2,,n(2)T2hyh(h=1,2)Var(yh)=1nni=1y2hi=h2T2hi=y2hih(3)ih,hihyhT2hi()iy1,y2:T2i=y21i1+y22i2(4)[7]2h=1y2ninHotelling,T2iT2(P,n-P)=T2(2,n-2),i=1,2,,n2(n-1)(n-2)T2iF(2,n-2)T2i(n-2)2(n-1)F(2,n-2)211a(0a1),iT2i(n-2)2(n-1)Fa(2,n-2),i();iT2iF(n-2)2(n-1)Fa(2,n-2),i()y21i1+y22i2En-22(n-1)Fa(2,n-2)¦Cay21i1+y22i2¦Ca(5)1,y1,y2i,y21i1+y22i2Ca,a;i,y21i1+y22i2FCa,a1211,,=0125,C0125=LCLT;=0105,C0105=UCLW,iT2i(n-2)2(n-1)F0125(2,n-2),()3:©1995-2005TsinghuaTongfangOpticalDiscCo.,Ltd.Allrightsreserved.(3)VI(t):VI(t)=Y21^1+Y22^2(6),Y1;Y2W=T2a(2,n-2)213,,:1VI(t)010LTTLTLWW+LWA+2VI(t)010510510817817LTLW,,:TW+A+(1);U0[3],A+,,;W+,,,W+,,,;T,,;LTU0,;U0LW,LTLWU0:LT=T2(P,n-1)=P(n-1)n-PF(P,n-P)(=0115)(7)LW=T2(P,n-1)=P(n-1)n-PF(P,n-P)(=0105)(8)U0=P(n-1)n-P-2(9),p=2,LW817,LT510,U0217,VI(t)2:214,LW(),,(),,,,,,(),,,,424320055©1995-2005TsinghuaTongfangOpticalDiscCo.,Ltd.Allrightsreserved.200312,sin-1x,,,(x-u)-1(x-u)FT2P()(=01050115),200312VI(t),33200312VI(t)111-2014413129-301142213-521031423-40109316-711421525-61142418-941561627-901185110-12018417210-1111426113-14114218212-1341067115-16018819214-1601628117-19619220217-1801599120-21114221219-20213410122-23017322221-23219811124-26111023224-25213612127-28114224226-2751352200312212,124,VI(t)1212,117-19226-2761925135,W+,,,,;,T,,12U0LCLT,,(38)5:©1995-2005TsinghuaTongfangOpticalDiscCo.,Ltd.Allrightsreserved.i)1,200190108085%2002315152632%,,0152102%1619787%,ii)1,019865,3(100-300),2(0-100),3,2,iii)1,1010513,13,,,,,,,,,,,[][1]Cios,K.J.,Pedryca,W.andSwiniarski,R.W.(1998),DataMiningMethodsforKnowledgediscovery[M].U.S.A,27-70.[2]Lebart,L.,Morineau,A.andWarwick,K.M(1984),MultivariaeDescriptiveStatisticalAnalysis[J].JohnWileyandSons,NewYork,109-145.[3]Pawlak,Z.(1982),RoughSets[J].InternationalJournalInformationComputerScience,11,341-356.[4]Johnson,R.A.andWichern,D.W.(1998),AppliedMultivariateStatisticalAnalysis[M].4thed,Prentice2Hall,Inc.[5](2003).DataMining[D].,49-57.(5)[][1].(DB440300/T20-2001)[S].,2001.[2](DB440300/T20-2001)[S].,2001.[3]Jackson,J.E.,T-SquareChartforBivariateData[J].PrincipalComponentsandFactorAnalysis:PartI-Principalcomponents,JournalofQualityTechnology12,201-213,1980.[4],,.[M].,1994.[5].SASSAS/STAT[M].,1997.[6].[M].,2003.[7]RichardA.Johnson,DeanW.Wichern.AppliedMultivariateStatisticalAnalysis[M].INTERNATIONALEDITION,1998.8324320055©1995-2005TsinghuaTongfangOpticalDiscCo.,Ltd.Allrightsreserved.